Background: Endoscopic retrograde cholangiopancreatography (ERCP) is often required to diagnose post-liver transplant (LT) biliary strictures. We evaluated the diagnostic accuracy of noninvasive laboratory and imaging tests in detecting post-LT biliary strictures.
Methods: Adult LT recipients who underwent ERCP between 2008 and 2015 were evaluated. Biliary strictures were diagnosed after blinded review of cholangiograms by three interventional endoscopists. The accuracy of liver enzymes, ultrasound, and MRI was determined using cholangiography as the reference standard. To evaluate the accuracy of change in liver enzymes, the difference between baseline and liver enzymes prior to ERCP (Δlab) was utilized.
Results: Biliary strictures were present on cholangiogram in 48 (58%) of 82 LT recipients meeting inclusion criteria. Baseline liver enzyme values did not differ significantly between patients with and without strictures. The optimal cutoffs for ΔALT, ΔAST, Δbilirubin, and Δalkaline phosphatase (AP) were determined to be 174 IU/L, 75 IU/L, 3.1 mg/dL, and 225 IU/L, respectively. ΔALT had a sensitivity of 100%, specificity 43%, and negative predictive value 100%. ΔAP had the highest specificity (53%) but modest sensitivity (69%) with a positive predictive value of 67%. Ultrasound had sensitivity of 29% and specificity of 69%, while MRI had sensitivity of 78% and specificity of 56%.
Discussion: The diagnostic accuracy of liver enzymes and imaging modalities is modest in detecting post-LT biliary strictures and cannot be used solely to identify patients needing further workup.
Keywords: Biliary stricture; Diagnostic accuracy; Endoscopic retrograde cholangiopancreatography; Liver transplant.